pipeline-flow-audit

To audit pipeline executions, the Skill compares actual events from decision and agent logs against expected stages defined in config files, generating a side-by-side comparison report with deviation analysis.

Verifying SDLC compliance involves reading a pipeline_run_id from Elasticsearch data, reconstructing the real stage sequence, and validating timing and responsibilities against defined config pipelines and workflows.

You generate a compliance report by providing a pipeline_run_id as the argument, which triggers an audit of planning, design, execution, and environment stages to surface deviations, retries, and escalations.

Yes, you can detect deviations and escalations by reconstructing actual stage sequences and agent activity from logs, then comparing them against expected SDLC workflows to identify anomalies.

Auditing SDLC pipeline flows requires config pipelines.yaml and workflows.yaml files to define expected stages, alongside decision and agent logs indexed in Elasticsearch containing the pipeline_run_id.

Pipeline audits show sequence and timing deviations when the actual events reconstructed from agent and decision logs fail to align with the expected stages and responsibilities defined in your config workflows.